Classifying physical motion contexts of a mobile device is useful for various applications. Such applications may include motion-aided geo-fencing, motion-aided Wi-Fi scan optimization, distracted pedestrian detection, health monitoring, etc. Common classifications may include walking, running, biking, fiddling, and being stationary, etc.
Motion contexts of a mobile device can be established through gathering and processing data received from sensors embedded in the mobile device. Mobile devices are now coming equipped with tri-axial accelerometers. A tri-axial accelerometer is a low-power sensor that can provide acceleration data representing acceleration components along orthogonal x, y, and z axes. A user's physical motion is transferred to a mobile device and the accelerometer embedded therein by either direct or indirect physical connection, such as by the user holding the mobile device in hand, or by the user keeping the mobile device in a pocket.